Re:In 27,280 coding agent runs, 64% of inference spend went ...
by SAI_Peregrinus on Tuesday November 17, 2026 @01:03PM
(#50535508)
When a task passes all tests, it's done. More tokens should be spent on failing runs, because most runs where tokens are used should fail. For the same reason that your lost keys are always in the last place you look for them.
[ Parent
| on HN ]
There are no comments on this comment yet.